1

A Green Approach of Improving Interface and Performance of Plant Fibre 1

Composites using Microcrystalline Cellulose 2 3

Subramani Pichandi, Sohel Rana*, Shama Parveen*, Raul Fangueiro 4 Centre for Textile Science and Technology (2C2T), School of Engineering, University of 5 Minho, Campus de Azurem, 4800-058 Guimaraes, Portugal 6 7

ABSTRACT 8

In contrast to the conventional methods of improving interface and performances of plant 9

fibre composites through fibre surface modification, this paper reports a novel approach 10

based on the dispersion of microcrystalline cellulose (MCC) in the composite’s matrix. MCC 11

was dispersed within the matrix of jute fibre reinforced epoxy composites to improve the 12

fibre/matrix interface as well as mechanical, dynamic-mechanical and thermal performances. 13

To develop these novel jute/epoxy/MCC hierarchical composites, MCC was first dispersed 14

within an epoxy resin using a short ultrasonication process (1h) and subsequently, the 15

MCC/epoxy suspensions were infused through jute fabrics using the vacuum infusion 16

technique and cured. Hierarchical composites by dispersing multi-walled carbon nanotubes 17

(MWCNTs) within the epoxy resin were also fabricated to compare their performance with 18

MCC based hierarchical composites. Interface (single fibre pull-out test), mechanical (tensile, 19

flexural, izod impact), thermal (thermogravimetric analysis) and dynamic mechanical 20

performances of the developed composites were thoroughly studied. It was observed that the 21

addition of MCC to the epoxy matrix led to a significant increase in the interfacial shear 22

strength (IFSS) between jute fibres and the epoxy matrix and consequently, resulted up to 23

18.4%, 21.5%, 28.3%, 67% and 49.5% improvements in the tensile strength, flexural 24

strength, impact energy, storage and loss moduli, respectively as compared to the neat 25

jute/epoxy composites. The above improvements achieved with MCC were significantly 26

higher as compared to the MWCNT based hierarchical composites developed using the same 27

technique. 28

29

Keywords: Micro-crystalline cellulose, jute, hierarchical composites, interface, mechanical 30

properties, dynamic mechanical performance. 31

32

*Corresponding author. Tel.: +351-910623763; Fax: +351-253 510 217. 33

Email address: soheliitd2005@gmail.com (S. Rana); parveenshama2011@gmail.com (S. 34

Parveen). 35

2

1. Introduction 36

In today’s world, there exists a growing awareness on the environmental problems of 37

different materials used in various industrial sectors. One of the main goals of research and 38

development worldwide is to develop materials or products which cause less harm to the 39

environment. As a consequence, the researchers have introduced plenty of plant based or 40

green materials most of which, however, lack in mechanical properties as compared to the 41

synthetic materials. A steady growth can be observed in the use of plant fibre based 42

composites in different industrial sectors including automobile, construction, transportation, 43

sports, etc. (Das, 2017). Besides environmental issues, plant fibres are becoming popular due 44

to their low cost, light weight as well as good specific strength and stiffness (Das, 2017). 45

Plant fibre composites also possess good thermal and acoustic insulation properties, excellent 46

electrical resistance, and high resistance to fracture (Parveen, Rana, & Fangueiro, 2017; Xie 47

et al., 2018). 48

49

However, a number of drawbacks such as poor mechanical performance, inferior fibre/matrix 50

interface, poor chemical and fire resistance and degradation problems are currently limiting 51

their full utilization (Parveen et al., 2017; Xie et al., 2018). Various chemical and physical 52

surface modification techniques have been extensively used to improve the inherent problems 53

with plant fibres and plant fibre composites (de Farias et al., 2017; Parveen et al., 2017; Xie 54

et al., 2018). The modification of plant fibre composites using nanomaterials is one of the 55

emerging technologies in this area. Plant fibre/fabric surfaces have been modified using 56

various nanomaterials to tailor the surface chemistry and morphology for using them in 57

composite materials (Blaker, Lee, & Bismarck, 2011; Rana, Parveen, Pichandi, & Fangueiro, 58

2018). Alternatively, different nanomaterials have also been introduced within the matrix of 59

plant fibre composites to improve their mechanical, thermal and dynamic mechanical 60

performances (Blaker et al., 2011; Rana et al., 2018). The incorporation of nanomaterials into 61

the plant fibre surface or into the matrix materials resulted in the development of hierarchical 62

or multi-scale composites, which were developed earlier mainly with the synthetic fibres 63

(Qian, Greenhalgh, Shaffer, & Bismarck, 2010; Rana, Alagirusamy, & Joshi, 2011; Wang, 64

Zhou, Wang, & Gao, 2016). 65

66

The hierarchical composites provide a new route of producing materials that show multi-67

functionality (i.e. improved mechanical performance, improved thermal stability and 68

conductivity, higher dynamic mechanical performance, lower co-efficient of thermal 69

3

expansion, higher electrical conductivity, self-sensing property, etc.) depending on the type 70

of nanomaterials used (Qian et al., 2010; Rana et al., 2011; Wang et al., 2016). CNTs have 71

been widely used to fabricate various synthetic fibre based hierarchical composites 72

possessing interesting properties (Qian et al., 2010; Rana et al., 2011; Wang et al., 2016). On 73

the contrary, only a limited number of research studies have been carried out on the CNT-74

plant fibre based hierarchical composites. Recently, jute fibres were coated with the CNT ink 75

and subsequently, the CNT coated jute fibres were impregnated with an epoxy resin to 76

produce hierarchical composites with strain and moisture sensing capabilities (Zhuang et al., 77

2011). 78

79

Due to environmental issues, health hazard and high cost of CNTs, currently the research 80

community became highly fascinated with the plant based green nanomaterials such as nano 81

cellulose. Nano and micro cellulose possess light-weight, high strength and stiffness, non-82

toxicity, bio-degradability, crystallinity and renewability, and these interesting properties 83

make them attractive for various industrial applications including packaging, cosmetic, food, 84

electronics, polymer composites, etc. (Ilyas, Sapuan, & Ishak, 2018; Ummartyotin, & 85

Pechyen, 2016; Zhao, et al., 2018). A number of studies demonstrated the use of nano 86

cellulose for developing plant fibre based hierarchical composites. For example, bacterial 87

cellulose (BC) was grown onto plant fibres (sisal and hemp) and the developed hierarchical 88

composites containing BC exhibited significantly improved fibre/matrix interface and 89

mechanical performance (Juntaro et al., 2008; Pommet et al., 2008). A coating of 90

nanocellulose onto jute fibres by Jabbar et al. (2017) led to improvements of tensile modulus, 91

flexural strength, flexural modulus, fracture toughness and storage modulus by 21%, 47%, 92

48%, 32% and 56%, respectively. The BC coating onto plant fibres also improved the 93

interface of plant fibres with cementitious matrices, prevented fibre mineralization and 94

improved the mechanical performance of composites (Mohammadkazemi, Doosthoseini, 95

Ganjian, & Azin, 2015). Besides surface application of nano/micro cellulose, their use in the 96

composite’s matrix can also improve the performance of plant fibre composites (Okubo, 97

Fujii, & Thostenson, 2009). For example, microfibrillated cellulose (MFC) was dispersed 98

within a PLA matrix to develop bamboo/PLA composites with an improved fracture energy, 99

resulting from the resistance of MFC to crack growth around the bamboo fibres (Okubo, 100

Fujii, & Thostenson, 2009). Looking at this promising result, in the present research micro-101

crystalline cellulose (MCC) was dispersed within an epoxy resin to improve the fibre/matrix 102

interface and the properties of jute/epoxy composites. Although plant fibre based hirarchical 103

4

composites have been previously developed using cellulose nanofibrils, BC and MFC, to the 104

best of authors’ knoweledge, this is the first attemp of developing hirarchical plant fibre 105

composites using MCC. One of the primary objectives of the present study was to develop 106

plant fibre composites with enhanced mechanical properties using inexpensive, readily 107

availble and green materials (jute fibres and MCC). The performances of these hirarchical 108

composites were compared with those fabricated using MWCNTs. The composites were 109

characterized thoroughly for fibre/matrix interface, tensile, flexural and impact properties, 110

thermal degradation and dynamic mechanical properties. 111

112

2. Materials and Methods 113

2.1 Raw Materials 114

Jute woven fabrics were purchased from Ribeiro Castro e Silva, Portugal. An epoxy resin 115

system (Biresin CR83 resin and Biresin CH83-2 hardener) was purchased from Sika®, 116

Germany. The resin and hardener were mixed in the ratio of 100:30 before preparation of 117

composites. MCC, Avicel® PH-101 grade, was purchased from Sigma Aldrich, Portugal. 118

MWCNT powder used in this work was purchased from Nanostructured & Amorphous 119

Materials, Inc., USA. Important properties of MCC, epoxy resin and MWCNTs are listed in 120

Table 1 and the scanning electron microscope (SEM) images of MCC and MWCNT are 121

presented in Fig.1. It can be observed that both MCC and MWCNT were highly 122

agglomerated in the powder. 123

124

Table 1 Properties of MCC, epoxy resin and MWCNT 125

Raw materials Properties Value

MCC*

Particle size ~50 µm

Moisture content 3 wt.%

Solid density 1.54 g/cm3

Particle shape Larger fibrous rods to smaller irregular cuboids

Epoxy resin**

Density (g/cm3) 1.15

Tensile strength (MPa) 122

Tensile modulus (GPa) 3.3

Elongation (%) 6.7

MWCNT*** Purity > 95%

5

Diameter Outer diameter: ≤ 8 nm, inner diameter: 2-5 nm

Length 10-30 µm

SSA > 500 m2/g

Bulk density 0.05 g/cm3

True density ~ 2.1 g/cm3

Melting point 3652-3679 °C

* Source: Sigma Aldrich, ** Source: http://sikaaxson.sika.com, ***Source: Nanostructured 126

& Amorphous Materials, Inc. 127

128

129

Fig.1. SEM image of MCC (a) and MWCNT (b) used in the present work 130

131

2.2. Dispersion of MCC and MWCNT in Epoxy Resin 132

The MCC powder (dried to remove moisture, if not stored in a moisture free container) was 133

added to the resin and then sonicated using an ultrasonicator (bath ultrasonicator, Sonica 134

Ultrasonicator 3200 S3, operated at 40 kHz frequency and 180W power) for 1h. Later, the 135

hardener was added to the resin/MCC mixture through gentle stirring. The final mixture was 136

used in the vacuum infusion process to produce hierarchical or multi-scale composites. The 137

dispersion technique used for MWCNT was slightly different from the process used for 138

MCC. MWCNTs show higher agglomeration tendency as compared MCC and therefore, a 139

mechanical stirring process at 1000 rpm was first introduced to mix MWCNTs with the resin, 140

prior to the sonication process. Next, ultrasonication of the resin/MWCNT mixture was 141

carried out for a longer duration (2h) for obtaining a good MWCNT dispersion. Finally, the 142

6

hardener was added to the resin/MWCNT mixture and stirred manually for 5 min before 143

preparation of the composites. 144

145

2.3 Fabrication of Hierarchical Composites 146

The vacuum infusion process was used to manufacture both MCC and MWCNT based 147

hierarchical composites. The composites were produced varying the concentration of MCC 148

and MWCNT, in order to investigate the effect of MCC or MWCNT concentration on the 149

composites’ performance. The weight percentages of MCC used were 1.0%, 2.0% and 3.0% 150

(with respect to the resin weight) and those for MWCNT were 0.5% and 1.0%. Lower 151

concentrations were used for MWCNT due its dispersion problem at higher concentrations. 152

For the comparison purpose, reference composite samples containing no MCC/MWCNT 153

were also manufactured. In all samples, 4 layers of woven jute fabrics were used. 154

155

2.4 Characterization of Mechanical Properties and Fibre/Matrix Interface 156

The tensile test of the composite samples was conducted according to ASTM D638-03 157

standard using a Universal Tensile Testing machine with a load cell of 50 kN and a crosshead 158

speed of 2 mm/min. For the measurement of elongation, an extensometer was employed 159

during testing. For each type, five samples were tested and the average tensile properties have 160

been reported. The flexural test was performed according to ASTM D790-3 standard. The 161

samples were subjected to a three-point flexural loading in a Universal Testing machine using 162

50 kN load cell, a span length of 60 mm and at a test speed of 2 mm/min. Five samples were 163

tested from each category and the average results have been reported. The fracture energy 164

was calculated from the area under the flexural load-elongation curves using Origin software. 165

Izod impact test was performed on the specimens according to ASTM D256 standard. 10 166

samples were tested from each type and the impact energy of the samples was expressed in 167

J/cm. Single fibre pull-out tests were carried out for jute/epoxy and jute/epoxy/MCC 168

hierarchical composites using the method discussed by Alderson et al. (2005). Interfacial 169

shear strength (IFSS) was calculated from pull-out tests in order to characterize the 170

fibre/matrix interface in jute/epoxy and hierarchical composites. Further, the morphology of 171

the fractured surfaces was studied using SEM (FEG-SEM, NOVA 200 Nano SEM, FEI, 172

acceleration voltage: 10kV, coating: 30 nm Au-P) to observe the adhesion between fibres and 173

the matrix in neat jute/epoxy and hierarchical composites. 174

175

7

2.5 Dynamic Mechanical and Thermogravimetric Analysis 176

The dynamic-mechanical properties of developed composites were studied using a dynamic 177

mechanical analyser (DMA, Hitachi DMA7100) at three-point bending mode using a sample 178

size of 50 x 14 x 1.5 mm (45 mm gauge length). Tests were performed at 1 Hz frequency and 179

the temperature was varied from 30° – 180° C at 3° C/min. In each type, 3 specimens were 180

tested and the average results have been reported. The thermogravimetric analysis (TGA) for 181

the produced composites was carried out using Hitachi STA700 instrument. The temperature 182

during TGA was varied from 30-900°C at 20° C/min rate in a nitrogen atmosphere. 183

184

3. Results and Discussion 185

3.1 Dispersion of MCC in Epoxy Resin 186

Fig. 2 shows the dispersion states of MCC and CNT in the epoxy suspensions, as observed 187

visually and through optical microscopy. 188

189

Fig. 2 Dispersion of MCC (a) and CNT (b) in epoxy resin at different concentrations 190

191

8

It can be noticed that the MCC suspensions were quite homogeneous without sedimentation. 192

The suspension with 1% MCC was transparent indicating a very good dispersion of MCC and 193

the transparency reduced with the increasing MCC concentration. The optical micrographs 194

also suggest that the MCC crystals were well dispersed without significant agglomeration at 195

1% MCC concentration, and as the concentration was increased further the agglomeration 196

also increased significantly (indicated by dotted circles). This finding is in agreement with the 197

previous studies reported in the literature (Saba et al., 2017). The visual observation of the 198

prepared CNT suspensions (Fig. 2b) also indicates homogeneous CNT dispersion without much 199

sedimentation, which was, however, difficult to observe clearly due to the black colour of the 200

suspensions. The optical micrographs showed the presence of CNT clusters, and upon 201

observations, it appears that their size increased with the increasing CNT concentration, as 202

also noticed in the previous studies (Frømyr, Hansen, & Olsen, 2012; Rana, Alagirusamy, & 203

Joshi, 2011; Xiao et al., 2018). 204

205

3.2 Mechanical Properties and Fibre/Matrix Interface 206

The tensile, flexural and impact properties of MCC based hierarchical composites along with 207

those of control samples are provided in Table 2. 208

209

Table 2 Tensile, flexural and impact properties of MCC based hierarchical composites 210

Samples σ

(MPa)

%

Increase ε (%)

σf

(MPa)

%

Increase

εf (%) FE

(N.mm)

IE

(J/cm)

Jute/epoxy 31.0 ±

1.5 --

0.41 ±

0.07

58.7±

5.3 --

2.76 ±

0.29

286.5

± 16.5

0.54 ±

0.03

Jute/epoxy/

1% MCC

36.7 ±

2.6 18.4

0.29 ±

1.02

71.4 ±

7.1 21.5

2.96 ±

0.14

461.3

± 29.1

0.62 ±

0.02

Jute/epoxy/

2% MCC

28.0 ±

6.3 -9.6

0.28 ±

0.05

66.8 ±

5.5 14.0

2.75 ±

0.37

460.3

± 47.0

0.69 ±

0.01

Jute/epoxy/

3% MCC

30.7 ±

3.9 -0.9

0.27 ±

0.02

61.8 ±

5.6 5.2

3.40 ±

0.39

468.0

± 22.7

0.62 ±

0.02

211

3.2.1 Tensile Properties and Fibre Matrix Interface 212

It can be observed that the addition of MCC up to 1% to jute/epoxy composites led to ~19% 213

improvement in the tensile strength (σ) of these composites. However, further increase in 214

9

MCC% led to decrease in the tensile strength because of the increased MCC agglomeration at 215

higher concentrations, as observed in Fig. 2. A similar trend has also been observed for 216

MWCNT. The addition of 0.5% MWCNT improved the tensile strength of jute/epoxy 217

composites to 34.5 MPa, achieving an improvement of tensile strength by 11.3%. The 218

hierarchical composites containing 1% MWCNT, however, exhibited a tensile strength of 219

32.1 MPa, which was only 3.6% higher as compared to the neat jute/epoxy composites and 220

much lower as compared to that achieved with 1% of MCC. Therefore, at the same 221

concentrations, the improvement was higher for MCC as compared to MWCNTs. As the 222

tensile strength is highly influenced by voids which are generated around CNT or MCC 223

agglomerates (as agglomerates are not impregnated by the resin properly during the vacuum 224

infusion process) (Xiao et al., 2018), the improvement in tensile strength was only observed 225

at lower MCC or CNT contents. This finding is in agreement with the previous studies which 226

showed increased MCC and CNT agglomeration at higher concentrations leading to lower 227

composite’s mechanical properties (Frømyr et al., 2012; Rana et al., 2011; Saba et al., 2017; 228

Xiao et al., 2018 ). The primary reason for the improvement in tensile strength of jute/epoxy 229

composites due to MCC addition was better compatibility of the MCC containing epoxy 230

matrix with jute fibres, resulting in a stronger interface. Previous researchers also believed 231

that the better interface was formed between jute fibres and epoxy matrices due to the nano 232

cellulose coating on the fibre surface; but they did not quantify the interfacial shear strength 233

(IFSS) (Jabbar et al., 2017). An improved interface was also found in plant fibre composites 234

through a BC coating on the plant fibre surface and these hierarchical composites showed 235

higher tensile strengths as compared to neat plant fibre composites (Juntaro et al., 2008; 236

Pommet et al., 2008). The results of pull-out tests on jute fibres with the neat epoxy matrix 237

and the epoxy matrix containing MCC are presented in Fig. 3 and the values of IFSS are 238

listed in Table 3. It is clear that MCC exhibited a strong influence on both initial and 239

maximum debonding loads of jute fibres from the epoxy matrix. Both increased with the 240

MCC addition and consequently, a significantly higher IFSS was achieved in case of the 241

epoxy matrix containing MCC (Table 3), leading to a higher tensile strength in case of 242

jute/epoxy/1% MCC composites. An improved IFSS between plant fibres and polymeric 243

matrices was previously achieved by coating BC/nanocellulose onto the plant fibre surface 244

(Juntaro et al., 2008; Pommet et al., 2008). The pull-out mechanism of jute fibres from the 245

epoxy matrix has been schematically presented in Fig. 3(b). 246

247

10

248

Fig. 3 Load-displacement curves of jute fibre during pull-out test (a), schematic of pull-out 249

mechanism (b) and flexural load-elongation curves (c) of jute/epoxy and hierarchical 250

composites 251

Table 3 Pull-out test results for neat epoxy and MCC/epoxy matrices 252

Samples Initial De-bonding

load, Fc (N)

Maximum De-bonding

load, Fmax (N)

Interfacial Shear Stress,

IFSS (MPa)

Epoxy 9.96 ± 1.1 35.2 ±2.4 0.65

Epoxy+ 1% MCC 12.35 ± 0.8 44.9 ± 3.6 0.83

253

The homogeneously dispersed MCC crystals formed a linkage between jute fibres and the 254

epoxy matrix in the hierarchical composites through chemical bonding. The formation of 255

covalent bonds between nanocellulose fibrils and epoxy molecules has been evidenced earlier 256

(Ansari, Lindh, Furo, Johansson, & Berglund, 2016) and therefore, could also be possible 257

between MCC and epoxy molecules. Further, jute fibres and MCC could form hydrogen 258

bonds due to presence of hydroxyl groups in their polymer backbones (Juntaro et al., 2008; 259

Pommet et al., 2008). Therefore, a strong jute fibre-MCC-epoxy network was formed in the 260

hierarchical composites, which resisted the debonding of jute fibres from the epoxy matrix. 261

11

The improved interface in case of hierarchical composites was further confirmed from the 262

fracture surface presented in Fig. 4. 263

264

Fig. 4 Fracture surface of neat jute/epoxy (a, b, c) and hierarchical composites (a1, b1, c1) 265

266

It can be clearly noticed from Fig. 4 (a) and 4 (a1) that the impregnation of jute fibres with 267

the epoxy matrix was much better in case of hierarchical composites. The presence of 268

individual and separated jute fibres could be observed in case of neat jute/epoxy composites 269

(Fig. 4b), whereas the fibres were strongly bonded to each other through the matrix in case of 270

hierarchical composites (Fig. 4b1). The surface of the jute fibres which were pulled out 271

during composite’s failure was quite smooth in neat jute/epoxy composites, as can be seen in 272

Fig. 4c. On the contrary, the fibre surface in case of hierarchical composites exhibited the 273

presence of surface damage and matrix fragments, indicating a strong adhesion with the 274

epoxy matrix (Fig. 4c1). This suggests that the interface of jute/epoxy composites changed 275

drastically due to the addition of MCC in the matrix. Similar improvement in the fibre/matrix 276

12

adhesion was previously noticed by dispersing CNTs and carbon nanofibres within the epoxy 277

matrix (Rana et al., 2011; Qian et al, 2010]. 278

279

This approach of improving interface by dispersing MCC within the composite’s matrix is 280

believed to be better as compared to the other chemical techniques (e.g. treatment with alkali, 281

silane, peroxides, permanganates, etc.) which are associated with considerable environmental 282

pollution and lead to the risk of fibre damage (de Farias et al., 2017; Kalia, Thakur, Celli, 283

Kiechel, & Schauer, 2013). The physical pre-treatments of plant fibres for improving 284

fibre/matrix interface like plasma and corona are clean and dry techniques, but the treated 285

fibres should be used immediately as the effects decay with time (de Farias et al., 2017; Kalia 286

et al., 2013). The recently reported nanocellulose coating technique on plant fibres to improve 287

interface needs an additional drying treatment after coating, which adds to an additional cost 288

and environmental issues (Jabbar et al., 2017). The BC coating technique (Juntaro et al., 289

2008; Pommet et al., 2008), on the other hand, is time consuming with a low production rate 290

and needs to be improved further for application at the industrial scale. On the contrary, the 291

approach presented here is a short, simple, dry and clean technique, which uses commercially 292

available and inexpensive green materials like MCC and can be readily opted by the current 293

industries. Further, the dispersion of MCC within the composite’s matrix, rather than coating 294

only on the fibre surface, will have more influence on the other important properties (e.g. 295

such as mechanical, thermal, dynamic mechanical, etc.) of composites. 296

297

In case of CNT based hierarchical composites, a higher tensile strength was achieved as 298

compared to the neat jute fibre/epoxy composites mainly owing to the improved fibre/matrix 299

interface and crack-bridging effect of CNTs (Borowski, Soliman, Kandil, & Taha, 2015; 300

Dikshit, Bhudolia, & Joshi, 2017; Rana et al., 2011). Due to the difference in co-efficient of 301

thermal expansion of epoxy and CNTs, a residual stress is generated during epoxy curing and 302

this residual stress increases the interface pressure and strengthens the fibre/matrix interface 303

(Rana et al., 2011). However, in the present research, the improvement of strength obtained 304

with MWCNTs was lower as compared to MCC. This could be mainly due to the CNT’s 305

agglomeration problem. MWCNTs could not be well dispersed using a 2h sonication process, 306

and a number of previous studies employed longer sonication treatments to ensure a 307

homogeneous dispersion (Frømyr et al., 2012). Nevertheless, it is interesting to note that the 308

breaking strain of the composites increased with CNT addition, whereas it decreased in case 309

of MCC based composites. This finding is in agreement with the previous studies (Haafiz et 310

13

al., 2013; Parveen, Rana, Fangueiro, & Paiva, 2017). The reinforcement of polymers or 311

cement with MCC reduced their breaking strains due to the stiffening effect of crystalline 312

MCC (Haafiz et al., 2013; Parveen et al., 2017). On the contrary, CNTs improve the breaking 313

strain and toughness of composites due to their high extensibility as well as crack-bridging 314

effects (Borowski et al., 2015; Dikshit et al., 2017; Rana et al., 2011). 315

316

3.2.2 Flexural Properties 317

The flexural strength (σf) and strain (εf) values of neat jute/epoxy and MCC reinforced 318

hierarchical composites are listed in Table 2 and the load-elongation curves of the composites 319

are presented in Fig. 3(c). The results showed that the addition of MCC to the neat jute/epoxy 320

composites enhanced both flexural strength and strain values. The highest improvement in 321

flexural strength (~22%) was noticed at 1% MCC, similar to the tensile strength and further 322

increase in the MCC content reduced the strength enhancement. As with the tensile property, 323

this could be attributed to the increased MCC agglomeration at higher concentrations, 324

resulting in reduced flexural strength of composites. However, the hierarchical composites 325

even with the highest content of MCC exhibited higher flexural strengths as compared to 326

control jute/epoxy composites. Also, it is noteworthy that the MCC addition to jute/epoxy 327

composites significantly improved their breaking strains and fracture energies (FE), as can be 328

seen from Table 2 and Fig. 3(c). Besides reinforcing effects, the main reason behind the 329

improvement of flexural properties due to MCC addition was the improved fibre/matrix 330

interface resulting in an efficient load transfer between fibres and the matrix in presence of 331

MCC. Previous studies also reported significant improvements in the flexural properties and 332

fracture energy in case of plant fibre based hierarchical composites (Jabbar et al., 2017 ; 333

Okubo et al., 2009). 334

335

3.2.3 Impact Energy 336

The impact energy (IE) of jute/epoxy and hierarchical composites are listed in Table 2. It can 337

be noticed that the impact energy improved up to 28.3% due to 2% MCC addition and further 338

increase in the MCC content reduced the impact energy improvement. The impact strength of 339

fibre reinforced composites depends on several factors including the type and quantity of 340

fibre, matrix and filler particles, fibre/matrix interface and also on the testing parameters (e.g. 341

type of test, whether notched or un-notched specimens, etc.). During impact testing, cracks 342

can travel through the polymer matrix as well as through the fibre/matrix interfacial regions 343

[30]. In case of hierarchical composites, the crack propagation could be slower due to crack 344

14

deflections at the uniformly dispersed strong MCC crystals, as also observed previously in 345

case of CNT reinforced polymers (Fiedler et al., 2006), and further due to the stronger 346

fibre/matrix interface. The strong interface favoured the deflection of propagating cracks 347

through the epoxy matrix leading to an enhanced impact energy absorption. In general, 348

impact properties of nano or micro cellulose reinforced polymers have been rarely studied in 349

the literature. A few earlier studies showed improvement in the impact strength due to nano 350

or micro cellulose addition (Mubarak, & Abdulsamad, 2018), while some researchers 351

observed the opposite trend (Kiziltas, Gardner, Han, Yang, 2010). The results of these 352

research studies were mainly influenced by the type of polymer and its compatibility with 353

MCC or nanocellulose. A good interface between MCC/nanocellulose and the polymer 354

matrix and homogeneous dispersion resulted in an improved impact strength (Mubarak, & 355

Abdulsamad, 2018), which however, deteriorated in case of poor interface with polymers or 356

inhomogeneous dispersion (Kiziltas et al., 2010). In the present case, a good compatibility 357

between MCC and the epoxy matrix and the resulting strong interface led to a significant 358

enhancement of impact energy in case of hierarchical composites. 359

360

3.3 Dynamic Mechanical and Degradation Behaviours 361

3.2.1 Dynamic Mechanical Properties 362

DMA was performed to characterize storage modulus (E'), loss modulus (E'') and Tanδ for 363

jute/epoxy and hierarchical composites. Storage modulus is a measure of stiffness of the 364

composites, whereas loss modulus presents the viscous response or un-recoverable oscillation 365

energy dissipation per cycle (Saba et al., 2017). Storage and loss moduli of neat jute/epoxy 366

and hierarchical composites are presented in Fig. 5(a) and (b). Tan δ curves of these 367

specimens are provided in Fig. 5(c). Tan δ presents the damping factor or energy dissipated at 368

the fibre/matrix interface of the composites (Saba et al., 2017). The values of storage and loss 369

moduli at 40°C and glass transition temperature (calculated from the peak of loss modulus 370

curves) are provided in Table 4. It is clear from Fig. 5 (a) that storage modulus of the 371

composites decreased with the increase in temperature as the composites passed from the 372

glassy state (with high stiffness) to the rubbery state with low stiffness. It can be observed 373

that the hierarchical composites presented significantly higher storage moduli in the glassy 374

state as compared to the neat jute/epoxy composites. 375

376

377

378

15

Table 4 Results of DMA and TGA analyses of jute/epoxy and hierarchical composites 379

Samples E'

(GPa)

Increase

in E' (%)

E''

(GPa)

Increase

E'' (%)

Tan δ

peak

height

Tg

(°C)

DTi

(°C)

DTm

(°C)

RW

(%)

Jute/epoxy 2.15 - 0.295 - 0.42 67.6 220.0 368.7 16.3

Jute/epoxy/

1% MCC

3.59 67.0 0.441 49.5 0.42 64.0 223.4 341.4 22.4

Jute/epoxy/

2% MCC

3.25 51.1 0.417 41.3 0.52 63.9 224.6 367.4 21.7

Jute/epoxy/

3% MCC

3.33 54.8 0.428 45.1 0.43 63.2 220.5 364.3 21.8

380

Storage modulus at 40°C increased by 67% in case of 1% MCC based composites and the 381

improvement reduced with further increase in the MCC content. The enhanced storage 382

modulus of hierarchical composites could be due to the enhanced stiffness of the epoxy 383

matrix owing to the presence of well dispersed MCC (Jabbar et al, 2017; Saba et al., 2017). 384

The mobility of epoxy molecules became restricted possibly due to the formation of stiffer 385

epoxy networks connected to high modulus cellulose crystals through chemical bonding 386

(Mubarak, & Abdulsamad, 2018). Improved jute/epoxy interfacial interactions in presence of 387

MCC could also improve the stiffness and therefore, storage modulus of the composites 388

(Jabbar et al, 2017; Saba et al., 2017). However, at the rubbery state, the molecular 389

movement of epoxy molecules became sufficiently high to suppress the stiffening effect of 390

cellulose crystals and as a result, storage modulus improved only marginally in the rubbery 391

state (Saba et al., 2017). Loss modulus of jute/epoxy composites also increased strongly in 392

case of hierarchical composites. The maximum improvement of loss modulus was noticed in 393

case of 1% MCC and further increase in the MCC content reduced loss modulus slightly, 394

similar to storage modulus. The presence of MCC improved the energy dissipation due to 395

increased interfacial friction between epoxy molecules and MCC as well as between jute 396

fibres and the epoxy matrix (Saba et al., 2017). Consequently, a strong increase in loss 397

modulus was noticed in case of hierarchical composites. The improvement of storage and loss 398

moduli of epoxy matrices through reinforcement with CNT and nano cellulose was 399

previously reported (Bhattacharyya et al., 2013; Saba et al., 2017) and the present results 400

16

showed the potential of MCC in improving the dynamic mechanical performance of 401

composites. 402

403

404

Fig. 5 Dynamic mechanical and degradation behaviour of neat jute/epoxy and hierarchical 405

composites: storage modulus (a), loss modulus (b), Tan δ (c), Cole-cole plot (d), thermal 406

degradation (e) and derivative weight loss curves (f) 407

408

It is also interesting to note that the enhancement of both storage and loss moduli decreased 409

with the MCC content. This is due to the fact that the increased MCC content increased the 410

number of MCC agglomerates in the matrix. MCC agglomerates could not be well 411

impregnated with the resin, resulting in generation of voids around them and reduced the 412

17

interactions with the epoxy molecules (Saba et al., 2017). The height of the Tan δ peak was 413

also higher for hierarchical composites indicating better damping characteristics of the 414

composites developed with MCC. From Table 4, it can be observed that Tg of the composites 415

reduced slightly due to MCC addition. Probably, MCC resulted in a reduced cross-link 416

density of the epoxy matrix, as previously observed for nano cellulose and CNT (Jabbar et al, 417

2017), and thereby, reduced Tg of the composites. 418

419

Cole-cole plots of the composite specimens are presented in Fig. 5(d). Cole-cole plot presents 420

the relationship between loss modulus (or Tan δ) and storage modulus of the composites and 421

can be interpreted to check the homogeneity of a polymeric system [20]. A smooth and semi-422

arc plot presents a homogeneous system, whereas imperfect or irregular shapes represent in-423

homogeneity or phase segregation or in-homogeneous dispersion of nano or micro fillers. 424

Although previous research studies demonstrated that the incorporation of nanocellulose led 425

to a deviation in Cole-cole plots of neat epoxy matrices (Saba et al., 2017), there was no 426

significant change observed in Cole-cole plots of neat jute/epoxy and hierarchical 427

composites. This suggests that the incorporation of MCC within the jute/epoxy composites 428

did not significantly change the homogeneity of the pure composite system. Although MCC 429

agglomeration at higher concentrations led to increased size of MCC agglomerates, it did not 430

result in segregation of MCC and matrix phases to have a significant influence on the 431

composite’s homogeneity and the shape of Cole-cole plots. 432

433

3.2.2 Degradation behaviour 434

TGA and differential thermogravimetry (DTG) curves of neat jute/epoxy and hierarchical 435

composites are presented in Fig. 5(e) and 5(f) and the results, i.e. the onset degradation 436

temperature (DTi), maximum degradation temperature (DTm) and residual weight (RW) have 437

been summarized in Table 4. It can be noticed that all specimens showed a similar 438

degradation behaviour in three stages. The first stage at temperature between 50°C to 150°C 439

was due to the evaporation of moisture from epoxy molecules, jute as well as MCC in case of 440

hierarchical composites. Following this, the second or main degradation stage was observed 441

between 220°C to 550°C with the degradation peak located at ~365°C. This wide degradation 442

temperature range at this stage was attributed to a number of degradation stages from 443

different components of the composites, including degradation of epoxy molecules between 444

300°C to 400°C (Saba et al., 2017), degradation of hemicellulose and lignin of jute between 445

280°C to 350°C (Kabir, Islam, & Wang, 2013), degradation of jute cellulose between 350°C 446

18

to 400°C (Kabir, Islam, & Wang, 2013) and degradation of MCC in case of hierarchical 447

composites between 275°C to 400°C (Trache et al., 2016). In the final degradation stage at 448

temperatures over 500°C, the weight loss was negligible and neat jute/epoxy composites 449

presented slightly lower residual weights as compared to the hierarchical composites. The 450

higher thermal stability of hierarchical composites at this final degradation stage could be due 451

to the increased char formation from MCC which acted as the insulating layer against the 452

thermal degradation of composites (Saba et al., 2017). 453

454

4. Comparison of MCC and CNT based hierarchical composites 455

The maximum improvements of mechanical and dynamic mechanical properties of MCC and 456

CNT based hierarchical composites have been compared in Table 5. It can be noticed that the 457

addition of MWCNT also resulted in significant improvements which, however, were lower 458

as compared to those obtained in case of MCC based composites. Although a longer 459

dispersion treatment (1h stirring + 2 h sonication) was used in case of MWCNT, a significant 460

amount of CNT agglomerates was present in the CNT based composites. This led to lower 461

property enhancements in case of MWCNT. MCC was easier to disperse within the epoxy 462

resin and a sonication treatment of 1h led to clear MCC/epoxy suspensions with very less 463

MCC agglomeration. These comparative results reflect the clear advantage of using MCC in 464

terms of processing and performance as compared to CNT for developing plant fibre based 465

hierarchical composites. 466

467

Table 5 Comparison of mechanical and thermal properties of MCC and CNT based 468

hierarchical composites 469

Improvement in properties (%) MCC based composites CNT based composites

Tensile strength 18.4 11.3

Impact strength 28.3 24.7

Storage modulus 67 44.6

Loss modulus 49.5 46.4

470

471

5. Conclusions 472

The addition of MCC to the matrix of jute fibre/epoxy composites was found to be highly 473

effective in developing hierarchical composites with superior fibre/matrix interface as well as 474

19

mechanical and thermal performances as compared to the neat jute fibre composites. The 475

presence of MCC within the epoxy resin improved the compatibility between jute fibres with 476

the epoxy resin and resulted in a significantly enhanced fibre/matrix IFSS. Due to MCC 477

addition, tensile, flexural and izod impact strengths of jute/epoxy composites improved by 478

18.4%, 21.5%, 28.3%, respectively resulting from the improved interface and reinforcing 479

effect of MCC crystals. The stiffening of epoxy molecules due to presence of MCC networks 480

led to strong improvements in both storage modulus (up to 67%) and damping factor or loss 481

modulus (up to 49.5%) of composites. In addition, the incorporation of MCC resulted in an 482

improved thermal stability of jute composites at the final degradation stage. The hierarchical 483

composites developed by dispersing MWCNTs within the matrix also showed significant 484

improvements in mechanical and dynamic mechanical properties, although lower as 485

compared to MCC based hierarchical composites. Therefore, MCC can be advantageously 486

used as an alternative green material to CNTs for developing plant fibre based hierarchical 487

composites. This research also explored a new green approach for improving the interface in 488

plant fibre composites using natural materials. In future, the studied approach will be used to 489

fabricate hierarchical composites using different types of plant fibres and polymeric matrices 490

and further optimized to apply for a wide range of plant fibre composites used in different 491

industrial applications. 492

493

References 494

Alderson, K. L., Simkins, V. R., Coenen, V. L., Davies, P. J., Alderson, A., & Evans, K. 495

E. (2005). How to make auxetic fibre reinforced composites. physica status solidi (b), 496

242, 509-518. 497

Ansari, F., Lindh, E. L., Furo, I., Johansson, M. K. G., & Berglund, L. A. (2016). 498

Interface tailoring through covalent hydroxyl-epoxy bonds improves hygromechanical 499

stability in nanocellulose materials. Composites Science and Technology, 134, 175-183. 500

Bhattacharyya, A., Rana, S., Parveen, S., Fangueiro, R., Ramasamy, A., & Joshi, M. 501

(2013). Mechanical and thermal transmission properties of carbon nanofiber‐dispersed 502

carbon/phenolic multiscale composites. Journal of Applied Polymer Science, 129, 2383-503

2392. 504

Blaker, J. J., Lee, K.-Y., & Bismarck, A. (2011). Hierarchical composites made entirely 505

from renewable resources. Journal of Biobased Materials and Bioenergy, 5, 1-16. 506

20

Borowski, E., Soliman, E., Kandil, U. F. & Taha, M. R. (2015). Interlaminar fracture 507

toughness of CFRP laminates incorporating multi-walled carbon nanotubes. Polymers, 7, 508

1020-1045. 509

Das, S. (2017). Mechanical properties of waste paper/jute fabric reinforced polyester resin 510

matrix hybrid composites. Carbohydrate polymers, 172, 60-67. 511

de Farias, J. G. G., Cavalcante, R. C., Canabarro, B.R., Viana, H. M., Scholz, S., & 512

Simão, R.A. (2017). Surface lignin removal on coir fibers by plasma treatment for 513

improved adhesion in thermoplastic starch composites. Carbohydrate polymers, 165, 429-514

436. 515

Dikshit, V., Bhudolia, S. K., & Joshi, S. C. (2017). Multiscale Polymer Composites: A 516

Review of the Interlaminar Fracture Toughness Improvement. Fibers, 5, 1-27. 517

Fiedler, B., Gojny, F. H., Wichmann, M. H. G., Nolte, M. C. M., & Schulte, K. (2006). 518

Fundamental aspects of nano-reinforced composites. Composites science and technology, 519

66, 3115-3125. 520

Frømyr, T. R., Hansen, F. K. & Olsen, T. (2012). The optimum dispersion of carbon 521

nanotubes for epoxy nanocomposites: evolution of the particle size distribution by 522

ultrasonic treatment. Journal of Nanotechnology, 2012, 1-14. 523

Haafiz, M. K., Hassan, M. A., Zakaria, Z., Inuwa, I. M., Islam, M. K., & Jawaid, M. 524

(2013). Properties of polylactic acid composites reinforced with oil palm biomass 525

microcrystalline cellulose. Carbohydrate polymers, 98, 139-145. 526

Ilyas, R. A., Sapuan, S. M., & Ishak, M. R. (2018). Isolation and characterization of 527

nanocrystalline cellulose from sugar palm fibres (Arenga Pinnata). Carbohydrate 528

polymers, 181, 1038-1051. 529

Jabbar, A., Militký, J., Wiener, J., Kale, B. M., Ali, U. & Rwawiire, S. (2017). 530

Nanocellulose coated woven jute/green epoxy composites: Characterization of 531

mechanical and dynamic mechanical behavior. Composite Structures, 161, 340-349. 532

Juntaro, J., Pommet, M., Kalinka, G., Mantalaris, A., Shaffer, M. S. P., & Bismarck, A. 533

(2008). Creating hierarchical structures in renewable composites by attaching bacterial 534

cellulose onto sisal fibers. Advanced materials, 20, 3122-3126. 535

Kabir, M. M., Islam, M., & Wang, H. (2013). Mechanical and thermal properties of jute 536

fibre reinforced composites. Journal of Multifunctional Composites, 1, 71-77. 537

Kalia, S., Thakur, K., Celli, A., Kiechel, M. A., & Schauer, C. A. (2013). Surface 538

modification of plant fibers using environment friendly methods for their application in 539

21

polymer composites, textile industry and antimicrobial activities: A review. Journal of 540

Environmental Chemical Engineering, 3, 97-112. 541

Kiziltas, A., Gardner, D. J., Han, Y., Yang, H.-S. (2010). Determining the mechanical 542

properties of microcrystalline cellulose (MCC)-filled PET-PTT blend composites. Wood 543

and fiber science, 42, 165-176. 544

Mohammadkazemi, F., Doosthoseini, K., Ganjian, E., & Azin, M. (2015). Manufacturing 545

of bacterial nano-cellulose reinforced fiber− cement composites. Construction and 546

Building Materials, 101, 958-964. 547

Mubarak, Y. A., & Abdulsamad, R. T. (2018). Effects of microcrystalline cellulose on the 548

mechanical properties of low-density polyethylene composites. Journal of Thermoplastic 549

Composite Materials, 1-15. 550

Okubo, K., Fujii, T., & Thostenson, E. T. (2009). Multi-scale hybrid biocomposite: 551

processing and mechanical characterization of bamboo fiber reinforced PLA with 552

microfibrillated cellulose. Composites Part A: Applied Science and Manufacturing, 40, 553

469-475. 554

Parveen, S., Rana, S., & Fangueiro, R. (2017). Macro-and nanodimensional plant fiber 555

reinforcements for cementitious composites. In H. Savastano Jr., J. Fiorelli, & S. F. 556

Santos (Eds.), Sustainable and Nonconventional Construction Materials using Inorganic 557

Bonded Fiber Composites (pp. 343-382). Amsterdam: Woodhead Publishing. 558

Parveen, S., Rana, S., Fangueiro, R., & Paiva, M. C. (2017). A novel approach of 559

developing micro crystalline cellulose reinforced cementitious composites with enhanced 560

microstructure and mechanical performance. Cement and Concrete Composites, 78, 146-561

161. 562

Pommet, M., Juntaro, J., Heng, J. Y. Y., Mantalaris, A., Lee, A. F., Wilson, K., Kalinka, 563

G., Shaffer, M. S. P., & Bismarck, A. (2008). Surface modification of plant fibers using 564

bacteria: depositing bacterial cellulose onto plant fibers to create hierarchical fiber 565

reinforced nanocomposites. Biomacromolecules, 9, 1643-1651. 566

Qian, H., Greenhalgh, E. S., Shaffer, M. S. P., & Bismarck, A. (2010). Carbon nanotube-567

based hierarchical composites: a review. Journal of Materials Chemistry, 20, 4751-4762. 568

Rana, S., Alagirusamy, R., & Joshi, M. (2011). Single-walled carbon nanotube 569

incorporated novel three phase carbon/epoxy composite with enhanced properties. 570

Journal of nanoscience and nanotechnology, 11, 7033-7036. 571

Rana, S., Parveen, S., Pichandi, S. & Fangueiro, R. (2018). Development and 572

Characterization of Microcrystalline Cellulose Based Novel Multi-scale Biocomposites. 573

22

In R. Fangueiro, & S. Rana (Eds.), Advances in Plant Fibre Composites (pp. 159-173). 574

Springer. 575

Rana, S., Ramasamy, A. & Joshi, M. (2011). Development of carbon nanofibre 576

incorporated three phase carbon/epoxy composites with enhanced mechanical, electrical 577

and thermal properties. Composites Part A: Applied Science and Manufacturing, 42, 439-578

445. 579

Saba, N., Safwan, A., Sanyang, M. L., Mohammad, F., Pervaiz, M., Jawaid, M., 580

Alothman, O. Y. & Sain, M. (2017). Thermal and dynamic mechanical properties of 581

cellulose nanofibers reinforced epoxy composites. International Journal of Biological 582

Macromolecules, 102, 822-828. 583

Trache, D., Hussin, M. H., Chuin, C. T. H., Sabar, S., Fazita, M. R. N., Taiwo, O. F., 584

Hassan, T. M., Haafiz, M. K. M. (2016). Microcrystalline cellulose: Isolation, 585

characterization and bio-composites application—A review. International journal of 586

biological macromolecules, 93, 789-804. 587

Ummartyotin, S., & Pechyen, C. (2016). Microcrystalline-cellulose and polypropylene 588

based composite: A simple, selective and effective material for microwavable packaging. 589

Carbohydrate polymers, 142, 133-140. 590

Wang, B., Zhou, G., Wang, L. & Gao, Y. (2016). Preparation and properties of 591

hierarchical composites based on carbon nanotubes‐coated glass babric preform. Polymer 592

Composites, 37, 979-986. 593

Xiao, C., Tan, Y., Yang, X., Xu, T., Wang, L. & Qi, Z. (2018). Mechanical properties and 594

strengthening mechanism of epoxy resin reinforced with nano-SiO2 particles and multi-595

walled carbon nanotubes. Chemical Physics Letters, 695, 34-43. 596

Xie, Q., Li, F., Li, J., Wang, L., Li, Y., Zhang, C., Xu, J. & Chen, S. (2018). A new 597

biodegradable sisal fiber–starch packing composite with nest structure. Carbohydrate 598

Polymers, 189, 56-64. 599

Zhao, T., Chen, Z., Lin, X., Ren, Z., Li, B., & Zhang, Y. (2018). Preparation and 600

characterization of microcrystalline cellulose (MCC) from tea waste. Carbohydrate 601

Polymers, 184, 164-170. 602

Zhuang, R.-C., Doan, T. T. L., Liu, J.-W., Zhang, J., Gao, S.-L., & Mäder, E. (2011). 603

Multi-functional multi-walled carbon nanotube-jute fibres and composites. Carbon, 49, 604

2683-2692. 605